CUAD合同抽取中的弃答错误与片段支持：可审计案例研究
原标题：Abstention Errors and Segment Support in CUAD: An Auditable Contract-Extraction Case Study
AI 摘要
该论文对CUAD合同条款抽取基准上的一个公开RoBERTa检查点进行了可复现的回顾性评估，修正了匹配错误并在62份训练合同上重新选择类别阈值。修正后的操作点以8.2%的CUAD精确率恢复了82.2%的参考片段，但在无标注答案的问题上出现45.4%的假阳性率。研究还发现41个类别中仅9个同时拥有至少30份正例和30份负例合同，两个类别完全没有负例合同，导致其无答案假阳性率无法定义。作者强调该结果仅描述单一检查点、解码器和阈值策略，不构成新的发布标准或生产就绪性。
正文节选
Abstention Errors and Segment Support in CUAD: An Auditable Contract-Extraction Case Study Abstract Contract-clause extraction benchmarks measure reference recovery, but interpreting a system for review assistance also requires measuring unnecessary output and the evidence available within categories. This paper presents a retrospective, reproducible evaluation of a fixed public RoBERTa checkpoint on CUAD’s published 102-contract test split. Candidate generation is preserved from an earlier froz